JP5587884B2 - Content caching in a radio access network (RAN) - Google Patents

Description

  This application claims the priority of US Provisional Application No. 61 / 086,521, filed August 6, 2008, which is hereby incorporated by reference.

  With the rapid growth of the Internet and the World Wide Web, there is a need for a substantially scalable data distribution solution for cable, DSL and other wired broadband networks. Even if several sites are mirrored or replicated to different geographical locations, the rapidly increasing data traffic cannot be handled. Content distribution networks (hereinafter also referred to as CDNs) have emerged to address the ever-increasing broadband subscriber and traffic scalability and performance issues. CDN uses various techniques including web caching to reduce bandwidth requirements, reduce server load, and improve user response time for cached content. Specifically, web caching refers to the storage of web document copies such as HTML pages, videos, images and other multimedia objects in a distributed cache, and subsequent requests for web content are cached when certain conditions are met. On behalf of CDN reduces the interactive round trip delay time of web browsing sessions by bringing content closer to the user. The CDN prefetches content before the actual request and stores it in the cache in order to increase the cache hit rate.

There are also wired providers that place web caches in their networks as shown in FIG. 1a to reduce internet bandwidth needs and increase subscriber browsing experience.
A content cache device or web cache that caches frequently viewed web pages, images, and multimedia content has traditionally been on the Internet to reduce transmission delays and download time for content that is heavily accessed. To place. Similarly, a web proxy / cache is also located at the corporate site to cache frequently used internet web content within the corporate network. Currently, these devices are used under certain restrictions in mobile radio networks.

  FIG. 1a shows an example network element of a wireless network commonly found today. A number of user devices 7 are connected to a local network medium such as DSL, cable or other Internet connection. A local DSL or cable backhaul 8 connects to the urban area network 9 via, for example, a DSLAM (DSL access multiplexer) or CMTS (cable modem terminal system) 11. The router 2 is used to move the packet via the Internet 12 according to the source and destination address of the packet. Server 14 hosts a website that includes the original content for the website. However, to save time and network traffic, a web cache 1 or other similar device that stores a copy of the original content is used. Thus, there can be one or more web caches 1 that provide requested data without burdening the server 14 throughout the Internet. In a large urban area, it is also common to introduce the cache server 1 into the urban area network 9.

The cache device may also be used in a mobile radio network, for example the 3G / UMTS network 20. Radio networks include a radio access network (hereinafter also referred to as RAN) and a core network (hereinafter also referred to as CN). FIG. 1b shows a typical wireless network.
A gateway GPRS service node (hereinafter also referred to as GGSN) 3 connects the mobile radio network to the IP core network. The GGSN 3 is a main component of a general purpose wireless packet service (hereinafter also referred to as GPRS) network. GGSN3 is a GPRS network, the Internet, X. Compatibility is provided with an external packet switching network such as a 25 network.

  Since GGSN3 hides the GPRS infrastructure from the external network, it appears as a router to the subnetwork from the external network. When the GGSN 3 receives the data addressed to the specific user, the GGSN 3 checks whether or not the user is active. If the user is active, the GGSN 3 distributes the data to a service GPRS support node (hereinafter also referred to as SGSN) 4 that supplies the service to the mobile user. However, if the mobile user is inactive, discard the data or initiate a paging procedure to locate and notify the mobile device. GGSN 3 routes these mobile-origin packets to the correct external network for data originating within the GPRS network.

  The GGSN 3 converts the GPRS packet sent from the RNC into an appropriate packet data protocol (PDP) format (for example, IP or X.25), and sends the converted packet on the corresponding packet data network. For incoming packets, the packet data protocol address is converted to the destination user's GSM address. Next, each address-converted packet is sent to the responsible RNC. To implement this function, GGSN 3 stores the user's current SGSN address and the user related profile in its location register. The GGSN 3 is responsible for IP address assignment and becomes a default router for the user's connected equipment (UE) 7. GGSN3 also performs an authentication function.

  The RNC delivers data packets to / from mobile stations within its geographic service area. RNC tasks include packet routing and delivery, mobility control (join / delete and location control), theoretical link control, authentication and charging functions. The RNC location register stores location information and user profiles of all GPRS users registered in the RNC.

  A radio network controller (hereinafter also referred to as RNC) 5 is a control element in the radio access network, and controls a node B6 connected to the radio network controller. The RNC 5 performs some of the radio resource control and mobility control functions, and is a point for encrypting user data before sending to / from the mobile. The RNC 5 is connected to an SGSN (Service GPRS Support Node) 4 in the packet switched core network.

Node B6 refers to a base transceiver base station (BTS) in the UMTS / 3GPP architecture. As is the case in all cellular systems, eg GSM, Node B (or BTS) 6 moves freely around Node B 6 using one or more radio frequency transmitters and one or more receivers. Communicate directly with user equipment.
The user equipment (hereinafter also referred to as UE) 7 includes all user equipment including a handset, a smartphone, and a computing device.

  For example, a radio access network (RAN) such as GSM / GPRS, 3G / UMTS / HSDPA / HSUPA, LTE, CDMA network has its own private network (PLMN), and a gateway device (GSM / GPRS, 3G / UMTS) / HSDPA / GGSUP in HSUPA and PDSN in CDMA) to the Internet / IP network. The content cache is typically located outside the RAN as shown in FIG. 1b. However, the content cache is not located in the RAN between the radio base station 6 and the GGSN 3 or PDSN (in the CDMA network).

  One reason is that although user application payloads are TCP / IP, these payloads are embedded within a radio access network protocol specific to a particular RAN. Thus, the application payload cannot be seen directly in the RAN during content aware caching and other optimizations. The RAN network 20 is arranged as a transmission network that transmits user IP traffic (owner IP traffic) via either ATM or IP transmission. Regardless of the transmission format, the RAN network transmits a user payload per service tunnel for each user. The service tunnel terminates in the PDSN or GGSN 3 that sends the owner IP traffic to the public IP network using IP forwarding rules. Thus, the conventional RAN network has non-content-awareness.

US Provisional Application No. 61 / 086,521

  Therefore, a cache device that operates within the RAN is beneficial. According to the cache device, access to content is further improved, and Internet traffic and transmission time are minimized. In addition, network elements in RAM are more localized, reducing capacity (throughput and concurrent users). This facilitates insertion of a device that has a low cache capacity and optimizes content-awareness. According to the network, distributed arrangement is facilitated and expandability is improved. There are benefits to methods and systems that allow intra-RAN caching.

  In accordance with the present invention, traffic is intercepted at standard interface points as defined by cellular / wireless networks (GSM / GPRS, 3G / UMTS / HSDPA / HSUPA, CDMA, WIMAX, LTE) and on each side of the intercept point A method is provided that emulates a protocol, extracts the user / application payload in an intercept packet, performs optimization, reencapsulates the same protocol, and delivers the content transparently. Optimization includes, but is not limited to, content cache, prediction and preemption of frequently used content, and content-aware transmission (TCP, UDP, RTP, HTTP, HTML, etc.) to reduce backhaul bandwidth And improved user experience. Additional embodiments of the present invention include injection of timely content (location-based, profile-based, history-based, or advertising content) based on information obtained during control plane protocol monitoring. The method of the present invention includes facilitation of cache and content delivery optimization by removing interface protocol layers at the intercepting interface, which includes a close packet inspection of user application packets, collection of business intelligence. There are enforcement of communication carrier definition policy control for protecting the RAN network of the communication carrier, user access right certification, prevention of access to unauthenticated contents (for example, parental control), and the like.

It is an example of arrangement | positioning of the content cache in each of the conventional wired network and a mobile telephone carrier network. It is an example of arrangement | positioning of the content cache in each of the conventional wired network and a mobile telephone carrier network. FIG. 2 is an illustration of 3GPP standard specifications network elements of a 3G / UMTS network of a cellular phone carrier and corresponding interfaces between these network elements. FIG. 3 is an exemplary diagram of a configuration in which a RAN cache (RANC) is arranged in an IuB interface between a Node B and an RNC in a 3GPP / UMTS network. It is an illustration figure of the structure which has arrange | positioned the RAN cache (RANC) in the IuPS interface between RNC and SGSN of a 3GPP / UMTS network. It is an illustration figure of the structure which has arrange | positioned the RAN cache (RANC) in the Gu interface between SGSN and GGSN of 3GPP / UMTS network. It is an illustration figure of the structure which has arrange | positioned RAN cache (RANC) in the S1 interface between eNodeB of a LTE / E-UTRAN network, and a MME / service gateway. FIG. 6 is an illustration of a control protocol layer in a RAN cache when operating as a double proxy between RCN and SGSN. FIG. 3 is an exemplary diagram of a user plane protocol that a RAN cache intercepts and extracts for each user traffic when placed in an IuPS interface of a 3GPP / UMTS network. FIG. 5 is an illustration of a control plane RAN cache operation between IuPS control plane traffic intercepts in the arrangement of FIG. 4. FIG. 6 is an illustration of a user equipment session initiated with a RAN cache that operates while another RAB is activated between two network nodes connected to a RAN cache. It is an illustration figure of the RAN cache which operate | moves as a double proxy by a user plane during the delivery of the cache content of a RAN cache, and a cache miss content via SGSN. FIG. 6 is an exemplary diagram of an environment in which a RAN cache is arranged on an IuPS interface that processes movement of a user equipment from one RNC range to another RNC range. FIG. 3 is a block diagram of a RAN cache according to one embodiment of the present invention.

  FIG. 2 shows a traditional 3G / UMTS network: user equipment 107, Node B (or base transceiver base station) 106, RNC (RNC or base base station controller) 105, SGSN (service GPRS support node) 104, A GGSN (gateway GPRS support node) 103 is included. FIG. 2 also shows the protocols used for communication between these various devices. For example, IuB 108 is a protocol used between the Node B 106 and the RNC 105. Similarly, IuPS is a protocol used between the RNC 105 and the SGSN 104. Gn 110 is used between SGSN 104 and GGSN 103. Gi 111 is an IP-based interface between the GGSN 103 and the Internet.

  FIGS. 3 to 5 illustrate an eavesdropping point into which a RAN cache (hereinafter also referred to as “RANC”) device can be inserted in a 3G / UMTS network. In FIG. 3, the RANC 112 is positioned between the Node B 106 and the RNC 105. In FIG. 4, the RANC 112 is positioned between the RNC 105 and the SGSN 104. In FIG. 5, the RANC 112 is positioned between the SGSN 104 and the GGSN 103. These diagrams are examples of deployment scenarios in 3G / UMTS networks, and do not show examples of deployment in other RAN networks, such as CDMA networks, etc., but the method described here is equally applicable to those networks .

  FIG. 13 shows a typical block diagram of RANC. The RANC 112 has two interface modules 201, each of which initiates the requested hardware signaling for the selected interface and associated software protocol. This interface protocol can be IuB, IuPS or Gn as shown in FIGS. Each interface module 201 is configured to transmit and receive via a selected interface. The received data is typically arranged in the storage element 202 which is a semiconductor storage element such as RAM, DRAM or equivalent technology. Data movement from the interface module to the storage element 202 and vice versa may be implemented using dedicated hardware, such as a DMA controller. Alternatively, the dedicated data movement processor can be used to actually move the data via the RANC 112. Information stored in the RANC 112 is processed according to the RAN specification.

  The processing is performed using a dedicated control logic or processing unit 203. The control logic / processing unit 203 may have its own local storage element 204, which stores instructions for execution and local status. The storage element can be RAM or DRAM. In addition, at least a portion of the storage element 204 can be non-volatile, such as ROM, flash ROM, hard disk, solid state disk. The control logic / processing unit 203 analyzes the received information using a known specification and protocol, and reads a packet at each protocol layer position. A mass storage element 205 adapted to hold cached information may also be included. In some embodiments, the cache storage devices can be semiconductor memories such as RAM or DRAM. In other embodiments, the cache storage device may be a rotating media such as a disk drive or other mass storage device. The control logic / processing unit can be physically initiated in various technologies. These technologies can be, for example, general purpose processors that execute instruction sets from internal and external storage devices.

  In other embodiments, the functions may be performed using a dedicated hardware device with embedded instructions or state machines. Throughout this specification, both “control logic” and “processing unit” shall refer to entities that are adapted to perform the set of functions.

  The RANC also stores software that can execute the functions described here. The software may be written in any public programming language, and the selection is not limited to the disclosure herein. Also, all applications and software described herein are computer-executable instructions that are stored on a computer-readable medium. For example, software and applications can be stored in read-only memory, rewritable memory, or embedded processing units. The specific computer executing the software has application dependency and is not limited by the present invention.

  FIG. 6 shows possible intercept points for RANC in a 4G / LTE network. In the LTE network, SGSN and GGSN are replaced by a mobility control entity (hereinafter also referred to as MME) / service gateway (hereinafter also referred to as SG) 113 and PDN gateway (hereinafter also referred to as PDN-GW) 114. Although not shown, in other embodiments, the MME and the service gateway are separate devices. Any embodiment is within the scope of the present invention. The MME 113 is a control node that is a key for the LTE access network. The MME 113 performs tracking and paging processing including retransmission in the UE (user equipment) 107 in the idle mode. The MME 113 is included in the bearer activation / deactivation process, and also performs service gateway selection at the time of initial connection of the UE 107 and handover within the LET including core network (hereinafter also referred to as CN) node relocation. . The MME 113 executes user authentication. The MME 113 confirms the camp-on authorization of the UE 107 on the service provider's general telephone network (hereinafter also referred to as PLMN), and implements the roaming restriction of the UE. The MME 113 is a termination point in the network for encryption / integrity protection for NAS signal communication, and handles security key control. Lawful intercept of signaling is also supported by MME 113.

  The service gateway (SG) routes and forwards user data packets, on the other hand, as a mobility anchor for the user plane during intra-eNode B handover, and as an anchor for mobility between LTE and other 3GPP technologies (S4 interface is terminated and traffic is relayed between the 2G / 3G system and the PDN-GW). The service gateway controls and stores the UE context, eg, IP bearer service parameters, network internal routing information.

  Thus, the MME serves as a control plane device, while the SG is a user plane device. Although these entities are physically separated, the interface to the MME is the S1 control plane and the interface to the SG is the S1-user plane. In an embodiment where each entity is physically joint, the interface is simply S1.

  The PDN gateway (PDN-GW) 114 provides a connection to the external packet data network of the UE 107 by serving as a traffic entry / exit point for the UE 107. The UE 107 can be simultaneously connected to one or more PDN-GWs 114 to access multiple PDNs. The PDN-GW 114 performs policy enforcement, packet filtering for each user, charging support, lawful intercept, and packet screening. Acting as an anchor for migration between 3GPP and non-3GPP technologies such as WiMAX and 3GPP2 (CDMA 1X and EvDO) is another important role of PDN-GW114.

In this environment, the RANC 112 can be inserted between the eNode B 106 and the MME 113 using the S1 interface 115 at both end positions. Alternatively, the RANC 112 can be inserted between the MME / service gateway and the PDN-GW 114 using the S5 interface.
When separating MME and SG, it is theoretically possible to place a RANC between eNodeB and MME in the control plane and between eNodeB and SG in the user plane using the S1 protocol. .

  Having described the various locations in the RAN network where the RANC can be located, its operation is described below. Although the protocol diagrams shown in FIGS. 7-11 are illustrated in the scenario of placing on the IuPS interface between the RNC and SGSN in the 3G / UMTS network, the method of the present invention is in a mobile network in the 3G / UMTS network. The present invention can also be applied to a case where it is arranged on another interface.

  FIG. 7 illustrates a control protocol layer in the RAN cache (RANC) when operating as a double proxy. The control plane is used for distributing information on users and various connections, for example, required quality of service (hereinafter also referred to as QOS), usage rules, relocation requests, and the like. In this example, the RANC appears to the SGSN proxy for the intercept point in the RNC 105 direction and to the RNC proxy for the intercept point in the SGSN 104 direction.

  FIG. 7 shows the upper part of the ATM transmission protocol layer as defined in the 3GPP protocol standard, and the lower part shows the IP transmission protocol layer. In this embodiment, the RANC stores RANAP (Radio Access Network Application Part) layer location IuP control packets and uses these packets to associate with a data plane tunnel (GPRS Tunneling Protocol User (or GTP-U) tunnel). To extract session information for each user. GTP-U is an IP-based relatively simple tunneling protocol that allows multiple tunnel formations between each endpoint set. A tunnel can be formed for each PDP context they have. Also, different QOS parameter sets can be specified for each tunnel.

  In operation, the software operates at each required level and analyzes the information. The remaining packet from which the protocol information related to the corresponding layer is removed is transferred to the next higher protocol layer. The process continues until the packet is completely decomposed. In the case of pass-through traffic, the packet is reproduced by adding protocol information as it passes down each layer. In other words, packet headers are re-added in the reverse order that they were removed, thus, L1 information is removed first in incoming packets and last in outgoing packets.

  FIG. 8 illustrates a user plane protocol, and a RAN cache (RANC) located on the IuPS interface in the 3GPP / UMTS network as shown in FIG. 4 is used to intercept each user traffic and extract those user traffic. The As can be understood from its name, the user plane is used to distribute user request data such as web pages and the like. In this figure, 3GPP specification IP and ATM transmission options are shown. The protocol names (L1, MAC, RLC, RRC, GTP-U, AAL5, ATM, etc.) in the figure are incorporated herein by reference in their entirety, and 3GPP TS25.401 of UTRAN Overall Description. Stipulated in

  FIG. 8 shows a RANC 112 that terminates the transmission protocol below the user IP layer in the IuPS stack on each interface of the RNC 105 and the SGSN 104. The RANC 112 also extracts the user payload and performs a transmission level or application level proxy + cache operation. These operations occur at level positions above the user IP level.

  The transmission level proxy operation includes, but is not limited to, transmission level connection termination such as TCP connection, application payload extraction, and application payload transfer in a new TCP connection at another interface. During payload transfer on a new connection, the payload is re-encapsulated using the same IuPS protocol at the second interface.

  Application proxy and cache operations include, but are not limited to, application protocols such as recognition of HTTP, RTMP, FTP, etc., recognition of object formats such as HTML, video objects, execution of application optimization, content cache operation, or both Execution. During the cache operation, the cache recognizes the user's request object and serves the request from the local content cache rather than forwarding it to the second interface.

  FIG. 9 illustrates the operation of the RAN cache (RANC) while intercepting IuPS control plane traffic in the arrangement shown in FIG. Each protocol layer shown in the figure is indicated by the specification of the 3GPP / UMTS standard. This figure shows a RANC that extracts GTP-U tunnel information 121 for each user session (per UE for a particular service type). The RANC then verifies the information 121, which can include, but is not limited to, user service type, device type, radio bearer, GTP-U / ID and GTP-U encapsulation type. Information 121 relates to each user session GTP-U traffic in the user plane (see FIGS. 9 and 11). Each GTP-U tunnel carries data traffic for a specific service flow of the UE in the user plane (see FIG. 8). Control plane monitoring in FIG. 9 facilitates identification of equivalent users, user equipment, QOS attributes for user sessions in the radio access network. 9 and 11 show optimizations that can be implemented to improve performance when using RANC. In this embodiment, an AAL5 (ATM conformance layer 5) packet is received on each interface. After reception, a local copy of the packet is saved. This local copy is then presented to higher layers for decoding and extraction of UE session information. At the same time, the received AAL5 packets are sent via path 127 to the second interface that transmits them.

  FIG. 10 illustrates a UE session started by RANC. In addition to monitoring control plane traffic as described with reference to FIG. 9, according to the present invention, it is possible to modify information elements in the control protocol and to insert a protocol packet for starting an additional session from an intercept point. The These operations are enabled by the RANC functioning as a double proxy at the intercept point positions of the control plane and the user plane as illustrated in FIGS. FIG. 10 shows a control protocol operation that causes the UE to initiate a new session to deliver content from a local cache for different service flows. In this scenario, the RANC multiplexes the control packets received from the SGSN for the RNC, and a new GTP-U tunnel is set in the user plane by the locally initiated session (secondary PDP) establishment process. Similarly, when receiving a control packet from the RNC, the RANC acts as a control plane SGSN, receives the control packet from the RNC, identifies a response to the locally initiated operation, and forwards the remaining packet to the SGSN. The RANC initiates a new session to improve QOS for pre-loading UE (eg well-viewed or user-configured content) or premium content (eg multimedia content requiring different QOS attributes). obtain. In other words, the RANC may also send and receive packets from the RNC that terminates at the RANC location while passing packets between the SGSN and the RNC.

  In FIG. 10, the RANC starts a second RAB (Radio Access Bearer) to distribute the stored content in the RANC, and another RAB is activated between the two network nodes connected to the RANC. Is shown. This figure shows user plane session traffic 122 that is locally initiated and multiplexed / demultiplexed with transit traffic at SM and MM locations for a specific UE.

  FIG. 11 illustrates the operation of the RAN cache (RANC) in the user plane as a double proxy while delivering the content cached in the RANC, the cache miss content from the SGSN, and the pass-through content. The user plane GTP-U traffic of the UE 121 is acquired using the equivalent session information in the control plane as shown in FIG. In some transactions, locally encapsulated data 123 is delivered from the RANC cache to the UE after encapsulation of the payload GTP-U protocol. Non-cached data or traffic 124 is forwarded from the SGSN interface to the RNC interface. A cache miss operation is also illustrated. In this operation, the RANC prefetches application data from the SGSN, fills the local cache, reencapsulates it with the corresponding user plane protocol, and forwards it to the RNC. In addition to sending data to the RNC, the RANC also stores the returned application data 125 in its content cache. The operations related to the cache data 123, the non-cache data 124, and the cache scheduled data 125 include processing specific to the additional protocol (TCP, UDP, HTTP, FTP, etc.) on the user IP layer (IP1, IP2, IP3). .

  FIG. 11 also shows the forwarding of another IP traffic 126 such as a non-GTP-U packet. These traffics are transferred by the RANC between the two interfaces connected to the RANC. Thus, in this figure, for each user GTP-U tunnel, the traffic forwarding part on the bearer plane IP layer (IP1, IP2, IP3) and the other part of the passing traffic at the position of the transmission IP layer 126 are shown. The RANC functions as a multi-protocol proxy while transferring bearer plane IP packets. For example, the RANC extracts TCP port 80 packets from the RNC and performs web cache and proxy operations. For other protocols such as FTP, RTP, RANC optimizes protocol-specific cache behavior, transmission (TCP / UDP) level or application layer before re-encapsulating user payload packets using corresponding interface protocol To do.

Hereinafter, the RANC operation will be described in detail with reference to FIGS.
First, for proper operation, the RANC should be theoretically invisible to surrounding devices. This is realized by acting as a proxy device. RANC intercepts control protocols such as IuB, IuPS or Gn and functions as a proxy. In other words, in the example of FIG. 4, the RANC 112 emulates the behavior of the SGSN 104 when the SGSN receives a control packet from the RNC 105. Similarly, the RANC 112 is visible to the RNC 105 while sending packets to the SGSN 104. RANC snoops packets from both RNC 105 and SGSN 104 for proxy function emulation and forwards them to another device. FIG. 7 shows the corresponding protocol layers in both the IP transmission embodiment and the ATM transmission embodiment between the RNC 105, the RANC 112, and the SGSN 104.

  The RANC 112 snoops the packet to identify the time of establishment of the data path tunnel (ie, connection of the PDP context) and determines the associated subscriber identifier. The RANC 112 may analyze a radio access bearer (RAB) establishment message within the RANAP protocol to identify the GTU-tunnel ID and the corresponding UE session. The process establishes a context between the data path tunnel (GTP-U tunnel) and the associated user.

  Alternatively, the RANC 112 does not learn the context from the IuPS control path, but rather monitors the new user plane tunnel (GTP-U tunnel) and the corresponding IP address (in the GTP-U tunnel). When recognizing a new GTP-U tunnel, the RANC 112 extracts a user IP address in the tunnel, searches for an external service plane network element such as a RADIUS server, and maps the IP address to a user identifier and a corresponding user profile. . Information such as whether the user is a prepaid or postpaid user and a rate plan (unlimited or restricted plan, type of authorized data service, etc.) may be obtained in this manner. In this case, access to the RADIUS server is required, but the amount of analysis context required from the transmission packet of the RANC 112 is minimized.

  UE information analyzed by snooping RAB establishment messages as described above or using a RADIUS server may also be used in a number of ways. For example, the RANC may obtain user profile information such as a price plan, user priority, authorization level (Internet general, carrier-specific on-deck content) using these techniques. Having obtained this information, the RANC prioritizes traffic and delivers data to the RNC during that time.

  For example, user rate plans may include monthly capacity limits (monthly megabytes / gigabytes), fair use when capacity limits are exceeded, best effort services, and the like. Once the user session is established, the RANC monitors the control plane protocol and determines that the specific UE is receiving a fair usage policy. Then, upon detecting a peak value approach situation due to congestion or increased data capacity to the RNC, the RANC restricts traffic to the UE's subject according to the fair usage policy. Traffic limiting methods include non-delivery from the local cache and / or optimization suspension for these user sessions.

In other embodiments, the carrier provides on-deck content as an additional service / rate plan. When the RANC determines that a specific UE has subscribed to on-deck content during session establishment, the RANC delivers the cached content from the on-deck site only when the user subscribes to that rate plan.
Another application may be provided by information analysis by monitoring the RANAP protocol. For example, by monitoring the RANAP protocol over an IuPS interface (see FIG. 7), the RANC can identify the user plane packet encapsulation type for each user session. This packet encapsulation type can be used to decode the user plane protocol for each session. Alternatively, the RANC may decode the tunnel packet type of each user GTP-U tunnel to fully decode the user plane packet.

  The RANC monitors the RANAP message, so that the top international mobile device identification number (IMEI) or mobile device number (MEI) and device (iphone (TM), blackberry (TM), laptop The type of computer, etc. may also be identified. In addition, or alternatively, the RANC can identify the web browser type (Internet Explorer, Safari, Fire Fox, etc.) of the user agent from the HTTP request in the user plane GTP-U packet received from the RNC.

  The device type thus determined is available for packet formatting or processing. For example, while delivering cached or non-cached (cache miss) content obtained from an Internet server, the RANC may perform device-specific content adaptation with UE type information. These adaptations include, but are not limited to, formatting for the screen size and selection of alternative files with different resolutions. In certain embodiments, a website may provide video at more than one resolution. The RANC may optimize the user experience by selecting a resolution that is best supported by the UE device based on the UE type determined by decapsulation and decoding of the RANAP message.

  As described above, the RANC can parse the RANAP message. By monitoring radio access bearer (hereinafter also referred to as RAB) assignment requests in the RANAP message, the RANC can identify the QOS parameters to use for a particular UE. Examples of QOS parameters include, but are not limited to, service type, maximum bit rate (MBR), guaranteed bit rate (GBR), and traffic processing priority. By obtaining these QOSs, the RANC can perform content optimization such as prioritizing audio streams during multimedia content distribution, for example. One example of application of QOS parameters by RANC is shown below. If the user establishes a RAB and the MBR parameter in the RAB Assignment Request message is 100 kbps specification, assume that the user has selected a high definition streaming video that requires 300 kbps. This user selection request is received by the RANC through a GTP-U tunnel corresponding to a specific RAB. RANC first determines if the multimedia object is in its cache or needs to be prefetched from the remote server. In any case, the RANC processes the protocol header and application specific content header (such as FLV header) and selects the content that requires 300 kbps. Since it is known that only 199 kbps is supported by the user equipment in RANC, only the audio portion of the stream is distributed to comply with the MBR parameter.

  An additional application of control plane information monitoring is UE connection area verification. In a certain wireless network arrangement, a certain area corresponds to a specific wireless sector. Since RANC intercepts all user plane traffic on the interface while correcting the control plane, it can identify all traffic targeted to a particular sector. The RANC can determine that a certain sector is near the occurrence of congestion based on all the identified traffic. Upon detecting sector congestion, the RANC may attempt to reduce congestion by limiting delivery traffic to peak users, limiting multimedia streams, or controlling traffic to certain types of devices (eg PC interface cards). .

  RANC can prioritize each user GTP-U traffic for the purpose of improving the quality of use for many users by using UE data in combination with QOS information as described above. Such prioritization includes, but is not limited to, peak user suppression and fair use policies during periods of congestion. For example, the RANC may detect that one or more devices make up the majority of traffic and throttle traffic to other devices. RANC may implement suppression for the device in question using certain algorithms.

  The RANC further decodes the IP-packet type, IP protocol type (TCP, UDP), and SRC / DST port number in each GTP-U tunnel, and application protocol type (for example, web / HTTP traffic), RTP traffic, FTP traffic. , RTMP, object type (eg, html, flv, .mp4, .mp3 file types). RANC may perform application specific decoding and optimization for each protocol type. Such optimization may include forming a cache for HTTP traffic as described above. For other protocols, such as FTP traffic, the RANC may perform a split TCP operation by separating the RANC-UE TCP connection from the RANC-Internet server TCP connection. FTP objects can also be cached by RANC using a cache and replacement policy for FTP objects. In another embodiment, RANC fulfills the re-transmission request from the UE local buffer for live streaming using RTP by maintaining the local transit buffer rather than forwarding the re-transmission request to the Internet server. .

  For each subscriber / GTP-tunnel, the RANC identifies the TCP packet and performs a TCP proxy operation. TCP proxy operations include, but are not limited to, maintaining another TCP connection to multiple UEs, maintaining a transit buffer, establishing multiple TCP connections while establishing a TCP connection in the direction of the core network requested by the server. Local retransmissions. Some TCP proxy operations may be known in the art, but are unique in that RANC needs to decapsulate and remove other interface protocols in order to provide these services. Other devices that perform TCP segmentation or proxy operations perform this operation during IP packet transmission. However, as shown in FIGS. 3-6, the RANC can operate using various interface protocols that embed user TCP / IP payloads within IP or non-IP transmissions, and can implement those TCP functions.

  As can be understood from the above-described embodiments, the RANC protocol allows the RANC to determine a specific action requested by a device on each side of the RANC. Based on this determination, the RANC may expand or modify the transmitted packet to better customize it for a specific UE. In another embodiment, RANC improves response time by caching some of the commonly used content and reduces the overall network volume.

RANC can maintain a cache of frequently accessed web pages, video clips, FTP files, etc. Their content cache can be shared among all users, thus cache replacement or refilling can be independent of the number of users. Thus, the latency is lowered and the quality of the user's feeling of accessing the top content is improved. Alternatively, the cache may be segmented such that each user's content occupies a percentage of the total cache. The mechanisms for determining what content to store in the RANC cache and how to segment the existing cache are known to those skilled in the art.
RANC may also maintain a history of frequently accessed content for each user. The RANC may maintain a minimum percentage of content for each user during content caching and replacement. Thus, caching can improve the overall quality of use for many users.

  In the user plane traffic received from the RNC, the RANC extracts the bearer IP packet in each user's GTP-U tunnel, identifies the protocol type and request information (eg URL information for HTTP traffic), and caches it locally. You can compare them with the content that was made. When the requested URL is found in the cache, the RANC returns a response, thus delivering the requested information. While returning those responses, the RANC emulates the SGSN and GGSN under conditions indistinguishable by the RNC. Therefore, the RANC forms a bearer IP packet and sends the packet into the corresponding GTP-U tunnel. RANC also emulates an FTP server for FTP traffic while returning a request file from the local cache. If the user request information does not exist in the cache, a request to emulate the RNC / IuPS process is reconstructed and forwarded to the core network (SGSN / GGSN).

The RANC may adjust the sequence number in each user service flow (GTP-U) when the sequence number option is used on the IuPS interface. For example, when the RANC delivers a cached object from its local cache, it is necessary to transfer a 100 GTP-U packet to the RNC. Since each GTP-U packet should have the highest sequence number, the corresponding sequence number in the GTP-U header is incremented. Since the object is not prefetched via SGGN, the sequence number used between RANC-RNC and SGSN-RNC is distinguished by adding 100. To compensate for this difference, the RANC may adjust the GTP-U sequence number for subsequent packets that are forwarded from the SGSN to the RNC for a specific GTP-U tunnel.
For protocols in the bearer IP plane where RANC does not provide performance support or caching, RANC receives the packet from one of the interfaces (RNC or SGSN) and re-adjusts the GTP-U sequence number if necessary, then the packet To another interface (SGSN or RNC).

  According to another aspect of the present invention, the content in the user plane is based on the information acquired from the control plane (such as the region and device type described above) and the content aware application processing in the user plane. Injected in a timely manner. The timely content is contextual and may be based on a user's access history, region, advertising content, and the like. For example, while processing an http request, the RANC processes both an http request and an http response. The RANC identifies the content type (such as an html page) while processing the http response. Upon timely content injection, the RANC may modify the html page to include additional URL links, additional html content or Java scripts, and so forth. Thus, the page received by the UE client includes not only the page content from the origin server but also the RANC designated content. Although the modification method of web page contents is conventionally known, in the present invention, these additional contents can be specified based on information acquired from the control plane and the user plane after decapsulating the interface protocol.

  Functions can be added by placing a RANC in the network. In one embodiment, the RANC is arranged between a mobile base station (Node B, BTS) and a mobile base station controller (RNC, BSC, WiMAX ASN gateway) as shown in FIG. In one such arrangement, the RANC is placed on the IuB interface in the 3G network. In this configuration, the RANC monitors the radio link control protocol exchanged between the Node B and the RNC. Thus, the RANC recognizes the radio link quality for each subscriber. RANC modifies content based on a wireless link to the mobile device. For example, the RANC may convert and prioritize specific types of content frames, such as only audio streams, depending on link quality, deliver only MPEG I frames, or discard packets for low quality links. Other actions that affect the content based on link quality may also be performed.

  The RANC of the specific embodiment may have a configuration for intercepting the BTS << RNC protocol (IuB interface protocol) as shown in FIG. In this embodiment, the RANC may monitor “Radio Bandwidth (OTA-BW)” by snooping CQI (Channel Quality Indicator) exchanged between BTS and RNC. The RANC then prioritizes and selects the optimal content for the UE using CQI. For example, the RANC may prioritize audio for video or web traffic for FTP traffic during delivery of multimedia stream content. The RANC may prioritize traffic based on the content format, such as character objects for image objects, during web page delivery. Thus, CQI utilization for traffic flow prioritization can be realized by the RANC. The decision regarding prioritization of traffic is in a specific application and is not limited by the present invention. Each of the above examples has no limiting purpose, but rather merely illustrates possible optimizations.

  While monitoring CQI as described above, the RANC may adjust the TCP congestion control window for a specific TCP session for the UE. The TCP congestion control window is an index of the number of protruding bytes at a specific time. This adjustment achieves, but is not limited to, maximum throughput to the UE (transverse traffic for all flows to a specific UE), reducing packet flow across all flows to the UE, and thus responding to new user input. It can reduce time and reduce the optimal throughput for all users while maintaining congestion at the RNC and fairness for active users.

  According to another aspect of the present invention, based on the radio bandwidth (OTA-BW) and round trip delay time (RTT) for TCP connection between the RANC and the UE, TCP optimization of user IP traffic of each UE is performed. Realized. In the above-described embodiment, the RANC arranged between the BTS and the RNC acquires the OTA-BW from the CQI. Alternatively, the RANC may obtain OTA-BW information by explicitly requesting those information from either the UE or the RNC. In another embodiment, the RANC obtains OTA-BW and RTT information by monitoring recent traffic to the UE or explicitly sending a protocol level or application level PING to the UE. Based on the evaluation of OTA-BW and RTT information, the RANC adjusts the initial TCP congestion window for maximum system throughput. Adjustment of the TCP congestion window based on throughput and RTT has been performed conventionally, but it is unique in that it obtains OTA-BW and RTT information by RANC.

  In radio access networks such as 3G, LTE, CDMA and WiMAX, mobility is one of the important aspects. Since the subscriber's handset or laptop can move from the coverage area of one cell base station (BTS or Node B) to the coverage area of another cell base station, the RANC can move on the basis of its location in the network. The problem (ie which interface to intercept) needs to be solved. Thus, when a specific mobile device distributed from a RANC in one location moves to another location, it is necessary to shift the content distribution context and related transfer status from the previous access RANC to the RANC in the new coverage area. There is. In the method of the present invention, active traffic via the RANC is continued in the new coverage area of the mobile radio environment, and the context is transferred between the two RANCs.

  FIG. 12 illustrates an environment in which a RAN cache for processing UE movement from one RNC range to another RNC range is arranged on the IuPS interface. The 3GPP standard defines movement and handover operations for UE movement processing from one RNC 105 (source RNC) to another RNC 105a (target RNC) within one SGSN or across two different SGSNs. 3GPP standard protocol and in particular 3GPP TS 25.410, which is hereby incorporated by reference in its entirety, defines a relocation procedure in which a source RNC 105 moves a UE's active session to another RNC 105a. . When the RANC 112 is placed between an RNC 105 and an SGSN 104 as shown in FIG. 12 on the IuPS interface, when the source RANC 112 recognizes the relocation of a specific UE, the source RANC is in service from its local cache or at the TCP / UDP level. Initiates context handover for content being transferred. Transmission traffic from the source RANC continues to pass through the target RANC.

  In order to support movement, each RANC communicates with its neighboring RANCs. Each RANC maintains a connected RNC identification number, a list of connected RANCs and RNCs. As described above, while monitoring the IuPS control protocol, the source RANC 112 connected to the source RNC 105 recognizes the relocation request of the target RNC and the identification number. The source RANC 112 determines a target RANC 112a to be connected to the target RNC 105a, and starts context transfer using the target RANC. The source RANC handles the relocation of the UE for which the source RANC is in content-aware operation to the target RANC by two basic operations. First, as shown in FIG. 12, the current RANC 112 transfers a UE context including user subscription, GTP-U tunnel information and other information to the target RANC 112a. Then, for ongoing transfers (eg, active TCP traffic), the source RANC 112 utilizes intra-RANC link 118 to direct traffic to and from the UE via its new coverage area (ie, new BTS 106a, target RNC 105a). And continue to send and receive. In the uplink direction (traffic received from the UE), the target RANC sends traffic for new flows (new TCP connection, DSN request, UDP request) instead of traffic for previous active flows (eg, TCP, ACks, RTP redistribution request, etc.). Identify. The target RANC forwards the previous active flow packet to the source RANC and processes the new flow traffic locally. In the downlink direction (UE traffic), the target RANC receives the previous active flow downlink packet from the source RANC and processes the new flow traffic locally. Thus, the source RANC 112 continues to supply cached content or any other TCP / UDP data for active flows. This step also includes the recognition of the new flow from the UE by the target RANC 112a and the simultaneous transfer of the previous active flow through the source RANC 112 at the same time.

5 ATM conformance layer 107 user equipment 115 S1 interface 118 intra-RANC link 121 GTP-U tunnel information 122 user plane session traffic 123 cache data 124 non-cache data 125 application data 126 IP traffic 127 path 201 interface module 202 storage element 203 control logic / Processing unit 204 Local storage element 205 Mass storage element

Claims (32)

A method for transmitting cache information in a radio access network (RAN), wherein the radio access network serves a plurality of users and includes a plurality of components, the method comprising:
a. Inserting a device between a first component and a second component in the radio access network, the device comprising a storage element, control logic, and two interface modules, wherein the device comprises the first and second components; Communicating with both of the two components;
b. Emulating one of the interface modules to make the second component appear to be the first component;
c. Emulating the other interface module to make the first component appear to be the second component;
d. Receiving communication from the first component for the second component, wherein the communication includes a plurality of control plane or user plane protocol layers;
e. Using the control logic of the device to identify when the data path tunnel is established and the associated subscriber identifier, thus establishing a context between the data path tunnel and the associated subscriber;
e-1. Analyzing the plurality of protocol layers using the control logic and thus determining the content of the communication;
f. Storing the content in the storage element;
g. Determining whether the communication is a request for content using the control logic, and if the request, determining whether the requested content is stored in the storage element;
h. Transmitting the stored content to the first component using the other interface module;
Including methods.   The method of claim 1, wherein the radio access network includes Node B, RNC, SGSN, GGSN, and the device communicates with the RNC and SGSN.   The method of claim 1, wherein the radio access network includes Node B, RNC, SGSN, GGSN, and the device communicates with the SGSN, GGSN.   The method of claim 1, wherein the radio access network comprises a Node B, an RNC, an SGSN, a GGSN, and the device communicates with the RNC, Node B.   The method of claim 1, wherein the plurality of layers includes a RANAP protocol layer, and wherein the control logic interprets the RANAP protocol.   The method of claim 1, wherein the radio access network includes eNode B, SG, MME, PDN-GW operating in an LTE network, and the device communicates with the eNode B, SG, MME.   The method of claim 1, wherein the radio access network includes eNode B, SG, MME, PDN-GW operating in an LTE network, and the device communicates with the SG, MME, PDN-GW.   The method of claim 1, wherein the device transmits the communication using the one interface module when the communication is not a request for content stored in a storage element. A method for distributing user specific information in a radio access network (RAN), wherein the radio access network serves a plurality of users and includes a plurality of components, the method comprising:
a. Inserting a device between a first component and a second component in the radio access network, the device comprising a storage element, control logic, and two interface modules, wherein the device comprises the first and second components; Communicating with both of the two components;
b. Emulating one of the interface modules to make the second component appear to be the first component;
c. Emulating the other interface module to make the first component appear to be the second component;
d. The control logic in the device is used to interpret communication from the first component to the second component to determine a user device for the communication and a specific parameter associated with the user, the communication having a plurality of controls Including a plain or user plane protocol layer;
e. Using the control logic to determine whether a response to the communication should be modified based on device intrinsic parameters;
f. Transmitting the communication to the second component using the one interface module;
g. Receiving a response from the second component using the one of the interface modules and modifying the response based on the determination;
h. Transmitting the response to the first component using the other interface module;
Including
Using the control logic of the device to identify when the data path tunnel is established and the associated subscriber identifier, thus establishing a context between the data path tunnel and the associated subscriber;
Analyzing the plurality of protocol layers using the control logic and thus determining the content of the communication;
Determining whether the communication is a request for content using the control logic, and determining if the requested content is stored in the storage element if the request is the request; Further comprising a method.   10. The method of claim 9, wherein the radio access network includes Node B, RNC, SGSN, GGSN, and the device communicates with the RNC and SGSN.   The method of claim 9, wherein the radio access network includes Node B, RNC, SGSN, GGSN, and the device communicates with the SGSN, GGSN.   10. The method of claim 9, wherein the radio access network includes Node B, RNC, SGSN, GGSN, and the device communicates with the RNC, Node B.   The method of claim 9, wherein the plurality of layers include a RANAP protocol layer and the control logic interprets the RANAP protocol.   10. The method of claim 9, wherein the radio access network includes eNode B, SG, MME, PDN-GW operating in an LTE network, and the device communicates with the eNode B, SG, MME.   10. The method of claim 9, wherein the radio access network includes eNode B, SG, MME, PDN-GW operating in an LTE network, and the device communicates with the SG, MME, PDN-GW. A network device adapted to operate on a radio access network, wherein each component of the radio access network communicates using a plurality of control planes or user plane protocols, the network device comprising the radio access network A first interface module and a second interface module adapted to communicate with the first component and the second component, respectively, wherein the first interface module appears to the first component relative to the second component Emulating the second interface module to appear to the second component relative to the first component;
A storage device;
Control logic interprets each of the plurality of protocols in the communication between the first component and the second component to identify the time of establishment of the data path tunnel and the associated subscriber identifier, and thus the data path tunnel and the association Establish a context with the subscriber to
The control logic is used to analyze the plurality of control plane or user plane protocol layers, thus determining the content of the communication, determining whether the communication is a request for content, and if the request Control logic for determining whether the requested content is stored in the storage element;
Including network equipment.   The control logic interprets the communication as an http request from the first component to a second component, and the control logic stores a response to the communication from the second component in the storage device. 16 network devices. The control logic is
Interpreting the communication from the first component to the second component, thus determining the user and content of the communication;
Determine whether the communication is a request for content, and if it is the request, determine whether the requested content is stored in the storage element;
The network device of claim 16, wherein the stored content is transmitted to the first component using the second interface module. The control logic is
Interpreting the communication from the first component to the second component, thus determining the user of the communication and the specific parameters associated with the user;
Determining whether to modify the response to the communication based on the specific parameter;
Transmitting the communication to the second component using the first interface module;
Modifying the response based on the determination;
The network device of claim 16, wherein the response is transmitted to the first component using the second interface module.   The plurality of protocols define a control plane and a user plane, and the control logic interprets communication between the first component and a second component to determine control plane parameters, and the control plane parameters are within the user plane. The network device according to claim 16, which is used in the above.   21. The network device of claim 20, wherein the control plane parameter is selected from the group consisting of radio access network congestion, user device information, user subscription plan, QOS attributes, and location information. A method for modifying communication to a user on a radio access network based on user-specific information, the radio access network including a plurality of components, the method comprising:
Inserting the device between a first component and a second component in the radio access network, the device comprising a storage element, control logic, and two interface modules, wherein the device comprises the first and second components; Said step of communicating with both of each component;
Emulating one of the interface modules to make the second component appear to be the first component;
Emulating the other interface module to make the first component appear to be the second component;
Interpreting communication from the first component to the second component using the control logic within the apparatus, the communication including a plurality of protocol layers, the plurality of protocols defining a control plane and a user plane; Steps,
Identifying per-user session information from the control plane, storing user attributes, and associating the user attributes with user sessions / tunnels in the user plane;
Modifying the communication for the user in the user plane location based on the stored user attribute and transmitting the modified information to the user;
Including
Using the control logic of the device to identify when the data path tunnel is established and the associated subscriber identifier, thus establishing a context between the data path tunnel and the associated subscriber;
Analyzing the plurality of protocol layers using the control logic and thus determining the content of the communication;
Determining whether the communication is a request for content using the control logic, and determining if the requested content is stored in the storage element if the request is the request; Further comprising a method.   23. The method of claim 22, wherein the control plane information is selected from the group consisting of radio access network congestion, user equipment information, user subscription plan, QOS attributes and location information. A method for delivering content to mobile users in a radio access network (RAN), wherein the radio access network includes a plurality of radio access network controllers (RANCs) and a cache device associated with each of the RANCs, The method inserts a device between a first component and a second component in a wireless network and emulates the first interface module of the device to make the first component visible to the second component; Emulating a second interface module of the device to make the first component appear to the second component;
Intercepting communications from the radio access network to the Internet using a first cache device, the communications including a plurality of protocol layers, the plurality of protocols defining a control plane and a user plane When,
Detecting the movement of the user by using the first cache device and monitoring the control plane;
Maintaining an association between the RANC and a corresponding cache device using the first cache device, thus determining a target cache device when a relocation request is detected in the control plane; ,
After the relocation request, performing a session handover from the first cache device to the target cache device;
Including
Using the control logic of the device to identify when the data path tunnel is established and the associated subscriber identifier, thus establishing a context between the data path tunnel and the associated subscriber;
Analyzing the plurality of protocol layers using the control logic and thus determining the content of the communication;
Determining whether the communication is a request for content using the control logic, and determining if the requested content is stored in the storage element if the request is the request; Further comprising a method.   25. The method of claim 24, further comprising the step of differentiating ongoing user plane traffic from new user plane traffic using the target cache device.   26. The method of claim 25, further comprising using the target cache device to deliver the ongoing user plane traffic to the first cache device.   27. The method of claim 26, further comprising using the target cache device for new user plane traffic.   26. The method of claim 25, wherein the user plane traffic includes a TCP connection.   The method of claim 1, wherein the communication is encapsulated within a tunnel.   The method of claim 1, wherein the first component and the second component utilize the same physical and logical protocols. Analysis of the plurality of protocol layers includes
Analyzing the request information of the first layer position,
Stripping the first layer related protocol information from the communication;
Repeat the analysis and stripping until the communication is completely decomposed,
The method of claim 1 further comprising the steps of: 32. The method of claim 31, further comprising reconstructing the communication using the control logic after complete disassembly of the communication.